DSpace https://erepo.uef.fi
Rinnakkaistallenteet Luonnontieteiden ja metsätieteiden tiedekunta
2017
Genome Sequencing and Population Genomic Analyses Provide Insights
into the Adaptive Landscape of Silver Birch
Salojärvi J
Springer Nature
info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
© Authors
CC BY http://creativecommons.org/licenses/by/4.0/
http://dx.doi.org/10.1038/ng.3862
https://erepo.uef.fi/handle/123456789/4943
Downloaded from University of Eastern Finland's eRepository
Forest ecosystems maintain a large share of Northern Hemisphere biodiversity. Boreal forests comprise roughly 30% of global forest area1 and contain the highest tree density among climate zones2. Moreover, boreal regions are undergoing extensive climate change.
Annual temperature increases exceeding 1.5 °C are projected to result in a warming of 4–11 °C by the end of this century, with little concomitant increase in precipitation1. At this pace, climate zones will shift northward at a greater speed than trees can migrate3. To understand how future populations of forest trees may respond to climate change, it is essential to uncover past and present signatures of molecular adaptation in their genomes. Silver birch, B. pendula, is a pioneer species in boreal forests of Eurasia. Flowering of the species
can be artificially accelerated4, giving it an advantage over other tree species with published genome sequences (such as poplar5, spruce6, and pine7) for the optimization of fiber and biomass production.
Here we sequenced 150 birch individuals and assembled a B. pendula reference genome from a fourth-generation inbred line, resulting in a high-quality assembly of 435 Mb that was linked to chromo- somes using a dense genetic map. We analyzed SNPs in the genomes of 80 birch individuals spanning most of the geographic range of B. pendula, as well as seven other members of Betulaceae. Population genomic analyses of these data provide insights into the deep-time evolution of the birch family and on recent natural selection acting on silver birch.
Genome sequencing and population genomic analyses provide insights into the adaptive landscape of silver birch
Jarkko Salojärvi
1,2,31, Olli-Pekka Smolander
3,31, Kaisa Nieminen
4, Sitaram Rajaraman
1,2, Omid Safronov
1,2, Pezhman Safdari
1,2, Airi Lamminmäki
1,2, Juha Immanen
1–3, Tianying Lan
5, Jaakko Tanskanen
2–4, Pasi Rastas
6,30, Ali Amiryousefi
1,2, Balamuralikrishna Jayaprakash
3,30, Juhana I Kammonen
3, Risto Hagqvist
7, Gugan Eswaran
1–3, Viivi Helena Ahonen
8,30, Juan Alonso Serra
1–3, Fred O Asiegbu
2,9, Juan de Dios Barajas-Lopez
10,
Daniel Blande
8, Olga Blokhina
1, Tiina Blomster
1–3, Suvi Broholm
2,11,30, Mikael Brosché
1,2,12, Fuqiang Cui
1,2,30, Chris Dardick
13, Sanna E Ehonen
1,2, Paula Elomaa
2,11, Sacha Escamez
14, Kurt V Fagerstedt
1,2, Hiroaki Fujii
10, Adrien Gauthier
1,2,30, Peter J Gollan
10, Pauliina Halimaa
8, Pekka I Heino
2,15, Kristiina Himanen
2,11,
Courtney Hollender
13, Saijaliisa Kangasjärvi
10, Leila Kauppinen
16, Colin T Kelleher
17, Sari Kontunen-Soppela
18, J Patrik Koskinen
3,30, Andriy Kovalchuk
2,9, Sirpa O Kärenlampi
8, Anna K Kärkönen
2,11,30, Kean-Jin Lim
2,11, Johanna Leppälä
1,2, Lee Macpherson
19, Juha Mikola
20, Katriina Mouhu
2,11, Ari Pekka Mähönen
1–3, Ülo Niinemets
21, Elina Oksanen
18, Kirk Overmyer
1,2, E Tapio Palva
2,15, Leila Pazouki
21, Ville Pennanen
2,15, Tuula Puhakainen
15,30, Péter Poczai
22, Boy J H M Possen
23,30, Matleena Punkkinen
10, Moona M Rahikainen
10, Matti Rousi
23, Raili Ruonala
3,30, Christiaan van der Schoot
24, Alexey Shapiguzov
1,2,25, Maija Sierla
1,2, Timo P Sipilä
1,2, Suvi Sutela
26,
Teemu H Teeri
2,11, Arja I Tervahauta
8, Aleksia Vaattovaara
1,2, Jorma Vahala
1,2, Lidia Vetchinnikova
27, Annikki Welling
1,30, Michael Wrzaczek
1,2, Enjun Xu
1,2,30, Lars G Paulin
3, Alan H Schulman
2–4, Martin Lascoux
28, Victor A Albert
5, Petri Auvinen
3, Ykä Helariutta
1–3,29& Jaakko Kangasjärvi
1,2Silver birch (Betula pendula) is a pioneer boreal tree that can be induced to flower within 1 year. Its rapid life cycle, small (440-Mb) genome, and advanced germplasm resources make birch an attractive model for forest biotechnology. We assembled and chromosomally anchored the nuclear genome of an inbred B. pendula individual. Gene duplicates from the paleohexaploid event were enriched for transcriptional regulation, whereas tandem duplicates were overrepresented by environmental responses.
Population resequencing of 80 individuals showed effective population size crashes at major points of climatic upheaval. Selective sweeps were enriched among polyploid duplicates encoding key developmental and physiological triggering functions, suggesting that local adaptation has tuned the timing of and cross-talk between fundamental plant processes. Variation around the tightly- linked light response genes PHYC and FRS10 correlated with latitude and longitude and temperature, and with precipitation for PHYC. Similar associations characterized the growth-promoting cytokinin response regulator ARR1, and the wood development genes KAK and MED5A.
A full list of affiliations appears at the end of the paper.
Received 24 January; accepted 12 April; published online 8 May 2017; doi:10.1038/ng.3862
OPEN
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
RESULTS
Genome assembly and gene duplication history
We constructed a hybrid nuclear genome assembly starting from 9× sequence coverage with Roche 454 technology. Assembled con- tigs were polished with Illumina paired-end data and connected and ordered using mate-pair libraries sequenced on both Illumina and SOLiD platforms, followed by further scaffolding and gap fill- ing with 30× coverage of PacBio reads longer than 6,000 bp. This resulted in a first assembly consisting of 5,642 scaffolds with an N50 of 240 kb (Supplementary Figs. 1 and 2 and Supplementary Tables 1 and 2). Further scaffolding with additional PacBio reads resulted in 3,474 (super)scaffolds with N50 value of 527.7 kb. A total of 391 Mb of scaffolds (89% of the estimated 440-Mb genome) was assembled into 14 chromosomal linkage groups via an ultra-high-density genomic linkage map consisting of 3.4 million markers (Fig. 1 and Supplementary Note). In addition to the nuclear genome, organel- lar genomes were assembled and annotated (Supplementary Note and Supplementary Figs. 3–6). Evidence for birch gene models was obtained by sequencing EST libraries from 12 different birch tissues or growth conditions, providing 18,951 transcripts with an average length of 1,683 bp, and by carrying out de novo assembly of RNA- seq reads, yielding 16,906 transcripts (Supplementary Figs. 7 and 8, Supplementary Tables 3 and 4, and Supplementary Note). We anno- tated the nuclear genome in two stages. After initial automated gene prediction, 3,192 genes were manually annotated and used to train gene predictors for a second round, yielding 28,153 high-quality gene models (Supplementary Figs. 9–11, Supplementary Tables 5–7, and Supplementary Note), of which 17,746 were supported by nearly full-length transcriptomic evidence.
Transposable elements (TEs) constituted 49.23%, and retro- transposons 30.60%, of the genome (Supplementary Note and Supplementary Table 8). Superfamily Gypsy and Copia retro- transposons were less common (8.5% and 2.3%, respectively), and contained fewer young (<50,000 years) elements than other plant genomes of similar size, whereas the nonautonomous TRIM group of retrotransposons was significantly more abundant8, at 6.4% ver- sus 1.26% (at most, in Pyrus, pear; P < 2.2 ×10−16, Grubbs test for one outlier). This suggests that the parasitic life cycle of TRIMs may attenuate replication of autonomous retrotransposons in B. pendula, thus limiting their contribution to genome size.
Syntenic alignment of the B. pendula genome with other eud- icots, including grapevine (Supplementary Figs. 12 and 13 and Supplementary Table 9), demonstrated that the birch genome has not undergone whole-genome duplications (WGDs) subsequent to divergence from these species (Supplementary Note). As such, the only internally duplicated blocks in the B. pendula genome date from the ancient gamma hexaploidy event at the base of core eudicots9,10.
Gene duplication and divergence is a major source of functional novelty in eukaryotic genomes, and in plants both polyploidy and tandem duplications have been implicated in the evolution of pheno- typic novelty11,12. Using self–self syntenic analysis, we separated the duplicated portion of the B. pendula genome into two pools: duplicate genes deriving from the ancient hexaploidy event, and those stemming from ongoing tandem (segmental) duplications (Supplementary Note).
Gene ontology (GO) functional enrichment analysis (Supplementary Tables 10–12) revealed that transcription factors (TFs) were strongly overrepresented among polyploid duplicates (Supplementary Table 12), which corresponds with theoretical and empirical results indicat- ing biased retention of highly interconnected genes following the duplication of entire functional modules13,14. This result appeared to hold for adaptive genome evolution in eudicots in general and
after independent WGDs, as we obtained highly similar functional enrichments in corresponding analyses of the Arabidopsis thaliana and poplar genomes15 (Supplementary Table 12), which have expe- rienced their own lineage-specific WGDs5,16 since diverging from a common ancestor with birch. In contrast to biased retention of modu- lar TF function after WGDs, tandemly duplicated genes in these three species15 were enriched for environmental responses and secondary metabolism, which, although distinctive by species, were also sig- nificantly overlapping (P < 2.2 × 10−16, Fisher’s test; Supplementary Note, Supplementary Fig. 14, and Supplementary Tables 11 and 12).
Tandemly expanded gene families shared by all three species were enriched for secondary metabolism, bacterial defense, hor- monal response, and hormone and nutrient transport. Adaptations possibly related to the arborescent habit were visible in convergent tandem expansions shared by B. pendula and Populus trichocarpa, including genes associated with fungal pathogen defense, cell wall biogenesis, and cellulose synthase activity (Supplementary Tables 11 and 12). Through evolutionary information stored within single genomes, these results suggest that whereas polyploid dupli- cates tend to diversify core processes in developmental and physi- ological regulation, tandem duplicates enhance the diversity of a plant’s environmental response capacity, which is in concurrence with previous studies15,17,18.
Population-level signatures of interspecies admixture
To examine recent B. pendula adaptation and to place the species into perspective within its parent clade, we sequenced the genomes of five other diploid birch species (Betula nana, Betula platyphylla, Betula populifolia, Betula occidentalis, and Betula lenta), the tetraploid birch Betula pubescens, two alder species from the related genus Alnus (Alnus incana and Alnus glutinosa), and B. pendula individuals origi- nating from 12 populations native to Ireland, Norway, Finland, and
TE density Gene density Bpe_Chr14
Bpe_Chr13
Bpe_Chr12
Bpe_Chr11 Bpe_Chr10
Bpe_Chr9
Bpe_Chr8
Bpe_Chr7 Bpe_Chr6
Bpe_Chr5 Bpe_Chr4
Bpe_Chr3 Bpe_Chr2 Bpe_Chr1
Sweep Tandem Syntenic
Figure 1 The pseudomolecule-level assembly of the silver birch genome shows an absence of WGD events since the ancient gamma hexaploidy.
Center, syntenic links between gene duplicates and triplicates dating from the hexaploid event. Ancient triplicate links still preserved in the modern genome are shown in red. Bpe_Chr, B. pendula chromosome.
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
Russia (Fig. 2a and Supplementary Tables 13 and 14). Additionally, eight ornamental varieties of B. pendula were included to scan for candidate gene mutations. SNPs were called using GATK19 for this collection of 89 individuals (Supplementary Table 15). Principal component analysis (PCA) of fourfold degenerate neutrally evolv- ing SNPs demonstrated that whereas most B. pendula individuals formed a single linear cline along PCA axis 2, another set consisting of eight B. pendula individuals was separated by PCA axis 1, following a trajectory suggestive of admixture from other birch species (Fig. 2b and Supplementary Fig. 15). Flow cytometric analysis of the latter set showed ploidy levels of four for two individuals where cambium material was available, and all atypical individuals showed high levels of heterozygosity (Supplementary Figs. 16 and 17), suggesting the possibility of novel polyploidies or admixture events with parents of polyploid origin.
Hybridization among birch species is well studied20,21. Explicit allele-sharing analyses with three-population F3 tests22 demon- strated traces of interspecific admixture within the main B. pendula population, suggesting that gene flow is ongoing and possibly bidirec- tional (Fig. 3 and Supplementary Table 16). Some B. pendula indi- viduals appeared to be highly admixed, including a few individuals from Punkaharju, Finland, where the main population was not admixed (Fig. 3c). The diploid species B. occidentalis and B. populifolia
displayed strong signatures of introgression from other species, including B. pendula (Fig. 3c and Supplementary Table 16). A phyl- ogeny of diploid birches and alders estimated from SNP data (Fig. 3a) placed B. lenta at the first split within the Betula genus, as suggested previously23, and supported by limited ribosomal DNA (rDNA) internal transcribed spacer (ITS) data24. Notably, when SNP called as a diploid, the tetraploid B. pubescens was placed in the same clade as B. lenta. Further analysis with three-population tests showed high levels of allele sharing between these two species (Supplementary Table 16), suggesting that a species closely related to B. lenta may be one of the diploid ancestors of B. pubescens. In an ITS-based phyl- ogeny24, B. pubescens was placed together with B. pendula, which suggests that it could be the second ancestor of the apparently allotetraploid B. pubescens.
For further analysis within B. pendula, we excluded the outlying individuals to reduce the confounding effects of interspecies admix- ture and polyploidy; this necessitated the removal of all Irish samples.
Analysis using ADMIXTURE showed very weak population structure with a split into two ancestral populations, roughly divided into east- ern and western groups (Fig. 2a and Supplementary Fig. 15), with some gene flow occurring between them, in Finland (Fig. 3c and Supplementary Table 16). This probably reflects allopatric division during the last Ice Age, followed by subsequent admixture when the
Germany n = 1 Ireland n = 4
Krasnoyarsk n = 4 Loppi n = 4
Posio n = 5
Punkaharju n = 23 Rovaniemi n = 5
Drøbak n = 4
Vehmersalmi n = 5
Voronezh n = 4
Yakutsk n = 4
Yekaterinburg n = 4
50 55 60 65 70
0 50 100
Longitude
Latitude
a
B. lenta B. nana (Fin)
B. nana B. occidentalis
B. platyphylla
B. populifolia B. pubescens A. glutinosa
A. incana
PC1 (35.4%)
PC2 (11.8%)
b
(Sco) Kittilä n = 5
Figure 2 Population genomics of silver birch and Betulaceae relatives. (a) Dispersion of 80 silver birches sampled from 12 sites across most of the geographic range of B. pendula. Populations are plotted with dots color-coded based on dispersion by latitude and longitude. ADMIXTURE analysis of SNP variation (superimposed with bar plots; center line, median; box, interquartile range; whiskers, 1.5× interquartile range; points, outliers) shows that Finland is a mixing zone between European (blue) and Asian (green) source populations. Samples from Ireland were highly admixed with other birch species and/or polyploid and were removed from the analysis. Source: OpenStreetMap contributors. (b) PCA shows clear separation between B. pendula populations and other sampled birch species (open circles). Eight B. pendula individuals (gray shading) were putative polyploids or interspecies admixed individuals. These included all Irish individuals and two individuals from Punkaharju, Finland. The main B. pendula population formed a cline along PC2 (purple shading). Fin: Finland, Sco: Scotland.
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
two populations rejoined after ice-sheet retreat, as has been suggested on the basis of chloroplast DNA evidence25. The small number of ancestral populations is probably due to the high degree of inter- breeding within birch populations; as a wind-pollinated species, birch pollen can spread more than 1,000 km26.
With the inclusion of ornamental cultivars, we sought to discover mutations in candidate genes that may account for their horticultur- ally interesting phenotypes. Among these were B. pendula ‘Youngii’, a weeping birch with a pendulous growth habit for which an in-frame stop codon was found in the birch AtLAZY1 ortholog (Supplementary Note and Supplementary Fig. 18). LAZY1 is a member of the IGT protein family27 and regulates tiller orientation in rice and maize as well as inflorescence branch angle in Arabidopsis28–30. It is thought to influence gravitropism through regulation of auxin transport and signaling28–30. Lateral organs in maize lazy1 mutants fail to grow ver- tically, giving rise to a phenotype similar to that observed in ‘Youngii’.
The stop codon in the birch LAZY1 ortholog could thus explain its weeping phenotype.
Population history shows ancient bottlenecks
For analyses of B. pendula effective population size (Ne) over time, we removed ornamental varieties, which do not have clear origin records, narrowing the analysis to 60 individuals. Within this set, linkage disequilibrium decay was slower (Supplementary Note and Supplementary Fig. 19), and nucleotide diversity (estimated to be 0.0088, Supplementary Table 17) was roughly 30% lower than in Populus tremula and Populus tremuloides31. The ancestral alleles for Betulaceae were reconstructed using the B. pendula reference genome and eight diploid Betula and Alnus species by estimating a phylogenetic tree and resolving ancestral states at nodes (Supplementary Note). The reconstruction was used to estimate the site frequency spectrum for the 60 B. pendula genomes using ANGSD32 and to generate a stairway plot33 elucidating Ne history over time (Fig. 4 , Supplementary Fig. 20,
and Supplementary Note). With a mutation rate estimate of 1 × 10−9 mutations per generation34 and a generation time of 20 years (Supplementary Note), the stairway plot revealed population bot- tlenecks over deep time that correspond roughly with known events of environmental upheaval (Fig. 4). An early Ne drop coincident with the great extinction event at the Cretaceous–Paleogene (K–Pg) boundary was clearly visible, followed by a rapid population expansion. Later Ne bottlenecks appeared during Eocene–Oligocene, mid-Miocene, and Pleistocene periods that correspond with other well-known episodes of environmental change35,36. The inferred history is largely supported by fossil evidence for Betulaceae23. Notably, the Ne dips we observed could be associated with cladogenetic events during Betulaceae his- tory, as alder–birch speciation occurred soon after the K–Pg event, 60 million years ago (Mya), and the mid-Miocene event ~14 Mya may have included the white-barked birches, for which there is fossil evidence from the late Miocene, 10 Mya23. In contrast, we observed no Ne bottlenecks during Holocene population history, for which a highly negative Tajima’s D of −1.82 would imply ongoing population expansion. Taken together, the B. pendula reference genome, rese- quenced individuals, and additional resequenced species provide an overview of the population genomic history for the entire Betulaceae clade over the past 65 million years.
Selective sweeps reveal coordinated local adaptation
We analyzed the same population of 60 resequenced silver birch indi- viduals for selective sweeps (Supplementary Note), where natural selection acting on a locus sweeps away variation across a genomic region surrounding the locus. Annotation of genes overlapping the sweep region or 2-kb flanking regions on either side indicated that some positive hits were probably artifacts resulting from recent inser- tions of chloroplast, mitochondrial, or TE DNA. Owing to their hap- loid nature at insertion, these horizontal transfers probably simulated homozygosity patterns reflective of selective sweeps. In total, 108
B. populifolia B. pendula B. platyphylla
B. occidentalis
B. nana (Fin) B. nana (Sco)
A. glutinosa A. incana
B. pubescens B. lenta 100
100 98
100 96
100 100
100
PC1 (14.2%)
PC2 (11.8%)
B. lenta B. nana (Fin) B. nana (Sco) B. platyphylla B. populifolia
B. pubescens Ireland Kittilä
Krasnoyarsk Kyynel−Youngii
Loppi Drøbak Posio Punkaharju admixed
Punkaharju Rovaniemi
A. incana A. glutinosa Vehmersalmi
Voronezh
Yakutsk Yekaterinburg
B. occidentalis
B. nana (Fin) B. lenta
B. nana (Sco)
B. platyphylla B. populifolia B. pubescens
Ireland Kittilä Krasnoyarsk
Kyynel−Youngii Loppi
Drøbak Posio
Punkaharju admixed Punkaharju Rovaniemi A. incana
A. glutinosa YakutskVoronezh Vehmersalmi
Yekaterinburg B. occidentalis
Source 1
Source 2
−30
−20
−10 0 10 Z.adj B. occidentalis
B. platyphylla
B. populifolia
B. pubescens
Punkaharju admixed
Punkaharju
Voronezh
a
b
c
Figure 3 Phylogenetic splits and admixtures among Betulaceae. (a) A phylogeny for Betula species, estimated from SNPs in neutrally evolving sites and noncoding sites 2 kb upstream and downstream of genes. Fin, Finland; Sco, Scotland. (b) PCA plot of 60 B. pendula individuals with low levels of admixture and known location information. Individuals diverging along PC2 are all from Punkaharju, Finland, indicating possible admixture from unknown source populations. (c) Heat maps for three-population F3 test statistics. Introgression is significant if the false discovery rate–adjusted Z-score (Z.adj; see supplementary Note) is significantly negative (adjusted Z-score < –1.96, purple).
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
genes near organellar or TE inserts were excluded, resulting in a final collection of 913 genes at or around which selection may have swept variation (Supplementary Table 18). This set was enriched for nontandem duplicates and single-copy genes. Tandemly duplicated genes did not show significant enrichment, suggesting that selection among tandemly expanded genes acts through a different process.
Regarding the age of the genes under selective sweeps, old syntenic orthologs were enriched for sweeps (P = 0.0018, Fisher’s test) whereas young birch-specific genes were significantly depleted of them (P < 2.2 × 10−16). Additionally, birch-specific nontandem genes were depleted of sweeps (P = 3.352 × 10−15), excluding a possible con- founding influence from tandem expansions. These results suggest that recent selective sweeps acted mostly on anciently diverged regu- latory components.
Although GO categorization has known pitfalls, it provides one of the best means to objectively characterize gene sets37. Exploratory functional enrichment analysis of genes in the sweep regions revealed three significantly enriched GO categories: transmembrane receptor protein tyrosine kinase signaling pathway (P = 1.24 × 10−5, Fisher’s test, Bonferroni adjusted); peptidyl-histidine phosphoryla- tion P = 3.91 × 10−5; and longitudinal axis specification (P = 0.00212).
These highlight known functions from model systems related to wood and fiber development, light sensing, embryogenesis, and reproduc- tive isolation. The first GO category includes 23 genes influenced by selective sweeps (Supplementary Table 19), most of which are phyloge- netically verified homologs of Arabidopsis genes encoding functionally characterized CLAVATA1-like receptor-like kinases (RLKs), including
BAM3, PXC2, PXC3, MOL1, MIK1, and MDIS1. In Arabidopsis, BAM3 controls leaf shape, size, and symmetry, as well as protophloem develop- ment38. The PXC genes are known to be involved in secondary cell wall formation in developing wood39. MORE LATERAL GROWTH (MOL1) is involved in cambium homeostasis, normally repressing secondary growth40. MDIS1-INTERACTING RECEPTOR LIKE KINASE1 (MIK1) is related to the PXC genes and also has a role in stem vascular devel- opment. MDIS1 forms a receptor complex with MIK1 and MIK2 that mediates the male perception of female chemoattractant LURE1 dur- ing fertilization in Arabidopsis41 and contributes to reproductive isola- tion between species. Transformation of AtMDIS1 to Capsella rubella partially broke down the interspecific reproductive isolation barrier41. Natural selection affecting birch MDIS1 therefore could suggest pos- sible involvement in determination of reproductive barriers between different birch species.
The second GO category highlighted by our exploratory analysis, peptidyl-histidine phosphorylation, includes nine phylogenetically verified orthologs of the phytochrome genes PHYA, PHYB, and PHYC, and genes encoding histidine kinases such as cytokinin receptors AHK2, AHK3, and AHK4 (CRE1), osmosensor AHK1, and ethylene receptors ERS1 and ETR2. The phytochromes are major mediators of red and far-red light responses that have vital roles in plant growth and reproduction42. The cytokinin and ethylene receptors control many key aspects of plant physiology and development, including acclimation to abiotic stress, shoot and root vascular development, flowering time, and longevity43,44.
The third GO category with putatively important functional enrich- ment, longitudinal axis specification, includes six phylogenetically verified genes including orthologs of Arabidopsis MONOPTEROS and GNL1, two homologs of TOADSTOOL 2, and two homologs of WRKY2.
MONOPTEROS and GNL1 operate in embryo and vascular develop- ment45,46; TOADSTOOL 2 operates in meristem maintenance47, and the Arabidopsis WRKY2 protein acts in zygote polarization in embryo development48. Additionally, WRKY2 has a role in pollen development49 and growth arrest induced by ABA during seed germination50. Candidate gene adaptation correlates with environment To assess whether selective sweeps were confounded by population structure, we performed redundancy analysis (RDA)51 on SNPs in the putative sweep regions by comparing the PCA eigenvectors from their SNP variation to overall population structure from PCA of whole-genome SNPs (Fig. 5 and Supplementary Note). In total, 423 of the 913 genes had a statistically significant (Benjamini-Hochberg adjusted P < 0.05 from permutation test) proportion between— 9%
and 100%— of their allelic variation explainable by overall B. pendula population structure (Supplementary Table 18), suggesting that these particular sweeps are at least partially confounded with drift processes.
After controlling for population structure, we identified genes show- ing clinal variation associated with general environmental variables such as temperature and precipitation. These restricting criteria resulted in a small subset of six genes with significant associations and intriguing molecular functions. These genes were verified by phy- logenetic analysis as orthologs of the A. thaliana genes SWEETIE, KAKTUS (KAK), ARABIDOPSIS RESPONSE REGULATOR 1 (ARR1), MED5A (encoding Mediator complex protein MED5A, also known as MED33A), PHYTOCHROME C (PHYC), and FAR1-RELATED SEQUENCE 10 (FRS10).
SWEETIE encodes a protein that may have a central role in sugar homeostasis52. In Arabidopsis, the sweetie mutant shows stunted growth, early senescence, flower sterility, and increased sugar levels.
In particular, the mutant has high levels of trehalose, a metabolite
1 × 105 1 × 106 1 × 107
0.77 1.82.58 14.8 34 66
Mya
Effective population size
Event Ple M E−Og K−Pg
Figure 4 Historical effective population size for silver birch beginning from 80 million years ago to present. Stairway plot showing that the B. pendula population has undergone bottlenecks during four known periods of major climate upheaval: the K–Pg (green) the Eocene-Oligocene (E–Og; blue), the mid-Miocene (M; red), and the Pleistocene (Ple; purple).
Data are median estimate from 200 bootstrap replicates (black line) and 95% confidence intervals (shading). Tick marks along the x axis show estimates for the Matuyama–Brunhes (0.77 Mya), Calabrian (1.8 Mya), and Gelasian (2.58 Mya) borders; the mid-Miocene disruption (14.5–14.8 Mya), and the E–Og (34 Mya) and K–Pg (66 Mya) events.
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
associated with signaling in plant interactions with microbes and her- bivorous insects, and in responses to cold and salinity. Additionally, sweetie shows altered expression for late embryogenesis-abundant (LEA) genes and many DREB2-type TFs. LEAs are anti-aggregation proteins that together with trehalose protect plant cells during the dehydration typical of abiotic stresses such as cold and drought53, whereas DREB2A and DREB2B are key transcription factors regulat- ing responses to dehydration and high-salinity stresses54. In birch, DREB TFs have been associated with cold acclimation and winter hardiness55. These connections may relate to the strong correlation silver birch SWEETIE shows with environmental variables.
KAK was identified originally as an endoreduplication repressor in Arabidopsis trichomes. However, the kak1 mutant shows increased C-values in etiolated hypocotyls completely lacking trichomes, sug- gesting a broader role in the control of endoreduplication56. KAK is also expressed in cambium, a secondary meristem that gives rise to both phloem and secondary xylem, where the gene has been suggested to have a role in defining the balance between xylem and phloem for- mation during vascular development57. In leaves, endoreduplication is associated with an increase in cell size and rapid growth, and also higher stress tolerance58,59. If KAK is indeed a general regulator of endoreduplication, its correlation with temperature may be of adap- tive significance to silver birch.
Together with cellulose and hemicellulose, lignin is an essential component of the secondary cell walls in structural fibers and water- conducting cells, determining their strength and rigidity. Lignin also interferes with the separation and breakdown of cellulose, hindering pulp and paper production and limiting the use of biomass crops for biofuel production. Attempts to reduce lignin production through genetic manipulation have so far resulted in plants with stunted growth and reduced yields60. MED5A was recently associated with lignin formation; the med5a mutation rescued the stunted growth, collapsed xylem vessels, and lignin deficiency phenotypes in the Arabidopsis phenylpropanoid pathway mutant ref8-1 (ref. 61). The double mutant med5a ref8-1 showed alleviated phenotypes, but cell wall properties were not restored to wild-type composition. Birch MED5A appears to be under positive selection, as several amino acid positions were significant by Bayes empirical Bayes positive selection analysis (Supplementary Note and Supplementary Table 20), its Tajima’s D value was in the lower 10% quantile for the B. pendula
genome, and its polymorphism correlated with latitude–longitude as well as temperature (Supplementary Table 18). A second com- ponent of the same mediator complex that appears among the putatively swept genes is MED12, which is involved in flowering time regulation in Arabidopsis62.
Cytokinin signaling is of pivotal importance for plant vascular development63,64; it is a major positive regulator of cambium activity and controls wood formation in the tree trunk65. We identified in our sweep list the birch ortholog of Arabidopsis transcription factor ARR1, a key regulator of Arabidopsis root meristem size66 that mediates the balance between cell division and differentiation by integrating auxin and cytokinin responses. ARR1 is involved in cold-induced inhibition of root growth and reduced auxin accumulation67, and it also con- trols Arabidopsis drought susceptibility68. This may explain the link to the geographic and temperature variables detected here. Variation in cytokinin signaling appears to have a large role in local adaption in silver birch, as the set of putatively swept genes also includes AHK2, AHK3, and AHK4, and their GO category (peptidyl-histidine phos- phorylation) was enriched, as described above.
Finally, the orthologs of PHYC and FRS10 are closely linked in the silver birch genome (Fig. 5), which is also the case in the grapevine and tomato genomes (Supplementary Fig. 21). PHYC and FRS10 act in red and far-red light sensing, shade avoidance, canopy density, temperature-dependent adaptation, and flowering time regulation69. PHYC was recently connected to temperature-specific regulation of the circadian clock70, and in Arabidopsis it is strongly linked to altitudinal71 and latitudinal–longitudinal70,72,73 clines in flowering time. In addition to PHYC, FRIGIDA and FLC explain a large pro- portion of flowering time variation in Arabidopsis71. Birch homologs of their Arabidopsis regulators FES, TXR7, and VEL1 were included in the selective sweep gene set. Because bud burst and initiation of flowering depend mainly on (night) temperature74, the function of PHYC in birch may be related to the photoperiodic control of inflores- cence initiation in the autumn, growth cessation, development of cold tolerance, and induction of senescence. PHYA and PHYB, encoding the other two main phytochromes, were also identified among the putatively swept genes, emphasizing the importance of light sensing for tree adaptation to varying environments. Phenotypic correlates of clinal variation in these and other genes remain obscure but are certainly worthy of more targeted analyses.
PC1 (62.1%)
PC2 (7.37%)
RDA1 (26.3%)
PC1 (50.5%)
Yekaterinburgkaterinkaterinb
Krasnoyarsk
Yakutsk
Yakutsk Krasnoyarsk
Jan Jul Dec
–37.1 0 25.3
Tmax
Coordinate on contig 243
CLR score
0 50 100 330 380 430
5 × 104 6 × 104 7 × 104 8 × 104 9 × 104 1 × 105 Sweep region
Gene under selection (mRNA region) Other gene (mRNA)
FRS10 MEF7 PHYC
Drøbak Kittilä
a b c
Punkaharju
Loppi Voronezh Vehmersalmi
Posio Rovaniemi
Figure 5 Selective sweep analysis reveals signatures of recent adaptation in the silver birch genome. (a) Putative sweep regions (yellow) around PHYC and FRS10 were obtained by accumulating and filtering composite likelihood ratio (CLR) statistics (red). Gene models are shown for each gene (purple), with the corresponding mRNA region highlighted in gray (or blue for genes under selection). (b) PCA plot based on SNPs 2 kb upstream and downstream of the sweep related to PHYC. (Sampling locations are color-coded as in Fig. 2). (c) RDA plot of the PHYC SNP region. Axis RDA1 is the direction that best correlates with principal components of maximum temperature (Tmax). In this case, Tmax explains 26.3% of the variation in the SNP data; the remaining variation is explained by PCA, where PC1 explains 50.5% (y axis). Inset, differences in yearly Tmax between Yakutsk and Krasnoyarsk, the locations most separated along RDA1.
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
DISCUSSION
Using the B. pendula reference genome and resequenced individu- als spanning the geographic range of silver birch, we were able to characterize genomic adaptations at several levels. First, we detected enrichments of TF functions that date to the core-eudicot crown radiation. Second, we uncovered a suite of gene duplicates involved in environmental responses that were not polyploidy derived but instead stem from ongoing tandem duplication processes. Such duplicates are generated by the same mechanisms as copy number variants (CNVs), which have come under intense recent study (particularly in animal genomes) as adaptive “tuning knobs” at the inter-population level75. In the case of silver birch, the tandem duplicates we observed might be taken as a sort of ‘species average’ that reflects former CNVs fixed by selection and neo- or subfunctionalization.
After several bottlenecks at well-known times of global envi- ronmental change, the effective population size of silver birch has increased over the past 1 million years. As expected for a wind- pollinated species with high pollen dispersal, population structure across the species range was particularly weak, which should greatly facilitate future GWAS efforts. Although we found evidence of ongoing gene flow between birch species, they were still clearly separated, and even if hybridization and introgression occurred, it did not blur their genetic distinctiveness. This contrasts with birch cytoplasmic markers, where evidence for allele sharing is common and species limits weak. Similar discordance between nuclear and cytoplasmic markers has been observed in other plant species76.
To identify recent selection in silver birch, we analyzed selective sweeps at the intraspecific level. These acted mostly on genes dat- ing from the ancient gamma hexaploidy event. Many of the genes located in the sweep regions were regulators and receptors that hold key positions in triggering developmental or physiological chains of events, suggesting that selection has acted during birch specia- tion by tuning the timing and cross-talk between different processes.
However, tandemly duplicated genes were not enriched for sweeps, and recently duplicated birch-specific genes were significantly depleted of sweeps. Altogether, these findings suggest an ‘exploration–
exploitation’ model for tandem duplicates in species evolution.
Exploration would occur through generation of novel tandem CNVs within populations; with lineage splitting and lowering of Ne, polymorphic tandems may become fixed by drift. Another alternative is fixation by selection through a process analogous to soft sweeps. As often reported in mammalian genome analyses77, multiple alleles at a locus can be swept to fixation in a ‘soft’ event that evades detection by ordinary criteria. In our example, tandem CNVs among individuals would comprise the alternative (exploratory) ‘allelic’ states, perhaps maintained in populations by balancing selection75. Multiple CNV states could then be simultaneously selected for when opportunistic and rapid (i.e., exploitative) responses to environmental change are required. Strong (‘hard’) selective sweep patterns, in contrast, involve selection for a single allelic state and may be more likely among core regulatory components that coordinate developmental timing and physiological cross-talk. Here, particularly when intertwined with population bottlenecks engendered by environmental upheaval, perhaps only genotypes that are unique and have proper timing of responses can be exploited.
We further uncovered candidate genes that show selection associ- ated with environmental responses and that are enriched for func- tions of practical relevance for forest biotechnology. Notably, several key components of cytokinin signaling, a major positive regulator of vascular cambium activity and radial tree-trunk growth65,78, show evidence of recent natural selection. Other examples are KAK and
MED5A, which can elicit pleiotropic growth and cell wall phenotypes in Arabidopsis. If orthologs of these genes function similarly in birch, this information could used for selective engineering of forest trees for rapid generation of biomass. Similarly, natural variation in pho- toperiod regulation might be used to understand and alter cambial activity–dormancy cycling and wood production.
METhODS
Methods, including statements of data availability and any associated accession codes and references, are available in the online version of the paper.
Note: Any Supplementary Information and Source Data files are available in the online version of the paper.
ACKNOWLEDGMENTS
We thank A. Korpijaakko, A. Korpilahti, S. Koskela, K. Lipponen, P. Laamanen, H. Kangas, M. Rantanen, and E.-M. Turkki for excellent assistance, and L. Schulman and L. Junikka (University of Helsinki Botanical Gardens) for providing birch species samples. Birch sequencing was supported by a Finnish Technology Development Agency (TEKES) grant to J.K., Y.H., and P.A. J.K. and Y.H. were supported by the Finnish Centre of Excellence in Molecular Biology of Primary Producers (Academy of Finland CoE program 2014-2019, decision 271832).
Y.H. was funded by the Gatsby Foundation and the European Research Council Advanced Investigator Grant SYMDEV. V.A.A. acknowledges support from the US National Science foundation (0922742 and 1442190). J.S. acknowledges a University of Helsinki 3-year grant. A.H.S. and J.T. acknowledge Academy of Finland decision (266430). EST libraries were created with TEKES funding to E.T.P.
AUTHOR CONTRIBUTIONS
J.K., P.A., and Y.H. conceived the study, and led the work together with J.S. and V.A.A. J.S. managed and coordinated all bioinformatics activities. L.G.P., A.L., and J.I. performed library construction and sequencing, and R.H., J.I., K.N., J.A.S., C.T.K., C.v.d.S., L.V., M.R., E.O., J.M., S.K.-S., G.E., A.P.M., and B.J.H.M.P.
participated in various aspects of biological sample collection, preparation, and quality control. O.-P.S. assembled the genome, with P.A. leading the work. O.-P.S.
assembled and P.S. and A.A. annotated the organellar genomes. P.R. carried out linkage mapping and anchored the genome into pseudochromosomes with J.S., O.-P.S., S.R., P.S., J.I.K., O.S. and A.A. J.I., R.R., L.K., A.W., T.P., P.I.H., L.G.P., E.T.P., Y.H., and J.K. produced EST libraries, which were analyzed by O.-P.S., J.S., and S.R.
J.I. and L.G.P. produced RNA-seq libraries, and O.-P.S., B.J., and S.R. analyzed the RNA sequencing data. J.S. and S.R. contributed to functional annotation, and P.J.G., H.F., J.V., T.H.T., S.E., A.V., A.K., M.W., S.O.K., A.G., A.S., V.P., M.M.R., M.B., M.P., S.K., O.S., T.B., K.M., K.H., J.P.K., S.S., F.O.A., P.H., F.C., K.V.F., K.-J.L., P.S., V.H.A., L.M., P.E., S.B., D.B., E.X., T.P.S., J.A.S., K.O., O.-P.S., O.B., A.K.K., J.I.K., M.S., A.I.T., J.L., J.d.D.B.-L., P.P., L.P., Ü.N., S.E.E., and K.N. participated in manual annotation, which was coordinated by P.S. and J.S. J.T. and A.H.S. annotated and analyzed the transposable elements. J.S., V.A.A., T.L., and S.R. performed comparative genomics analyses. J.S., V.A.A., O.S., and M.L. analyzed population genomics data. J.S., V.A.A., and T.L. prepared figures. J.S. and V.A.A. wrote the paper, with input from J.K., M.L., K.N., Y.H., P.A., A.H.S., O.-P.S., C.D., and J.I. J.S., O.-P.S., V.A.A., L.G.P., S.R., A.L., R.H., J.I., O.S., K.N., M.R., A.H.S., P.R., P.S., A.A., and G.E. provided text for the supplement. K.N. and S.R. had equivalent overall roles in the project. All authors approved the final version of the manuscript.
COMPETING FINANCIAL INTERESTS The authors declare no competing financial interests.
Reprints and permissions information is available online at http://www.nature.com/
reprints/index.html. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This work is licensed under a Creative Commons Attribution 4.0 International (CC BY 4.0) license. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
1. Gauthier, S., Bernier, P., Kuuluvainen, T., Shvidenko, A.Z. & Schepaschenko, D.G.
Boreal forest health and global change. Science 349, 819–822 (2015).
2. Crowther, T.W. et al. Mapping tree density at a global scale. Nature 525, 201–205 (2015).
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
3. McKenney, D.W., Pedlar, J.H., Lawrence, K., Campbell, K. & Hutchinson, M.F.
Potential impacts of climate change on the distribution of North American trees.
Bioscience 57, 939–948 (2007).
4. Longman, K.A. & Wareing, P.F. Early induction of flowering in birch seedlings.
Nature 184, 2037–2038 (1959).
5. Tuskan, G.A. et al. The genome of black cottonwood, Populus trichocarpa (Torr. &
Gray). Science 313, 1596–1604 (2006).
6. Nystedt, B. et al. The Norway spruce genome sequence and conifer genome evolution. Nature 497, 579–584 (2013).
7. Neale, D.B. et al. Decoding the massive genome of loblolly pine using haploid DNA and novel assembly strategies. Genome Biol. 15, R59 (2014).
8. Gao, D., Li, Y., Kim, K.D., Abernathy, B. & Jackson, S.A. Landscape and evolutionary dynamics of terminal repeat retrotransposons in miniature in plant genomes.
Genome Biol. 17, 7 (2016).
9. Jaillon, O. et al. The grapevine genome sequence suggests ancestral hexaploidization in major angiosperm phyla. Nature 449, 463–467 (2007).
10. Jiao, Y. et al. A genome triplication associated with early diversification of the core eudicots. Genome Biol. 13, R3 (2012).
11. Jiao, Y. & Paterson, A.H. Polyploidy-associated genome modifications during land plant evolution. Phil. Trans. R. Soc. Lond. B 369, 20130355 (2014).
12. Flagel, L.E. & Wendel, J.F. Gene duplication and evolutionary novelty in plants.
New Phytol. 183, 557–564 (2009).
13. Papp, B., Pál, C. & Hurst, L.D. Dosage sensitivity and the evolution of gene families in yeast. Nature 424, 194–197 (2003).
14. Freeling, M. & Thomas, B.C. Gene-balanced duplications, like tetraploidy, provide predictable drive to increase morphological complexity. Genome Res. 16, 805–814 (2006).
15. Rodgers-Melnick, E. et al. Contrasting patterns of evolution following whole genome versus tandem duplication events in Populus. Genome Res. 22, 95–105 (2012).
16. Arabidopsis Genome Initiative. Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408, 796–815 (2000).
17. Hanada, K., Zou, C., Lehti-Shiu, M.D., Shinozaki, K. & Shiu, S.-H. Importance of lineage-specific expansion of plant tandem duplicates in the adaptive response to environmental stimuli. Plant Physiol. 148, 993–1003 (2008).
18. Chae, L., Kim, T., Nilo-Poyanco, R. & Rhee, S.Y. Genomic signatures of specialized metabolism in plants. Science 344, 510–513 (2014).
19. DePristo, M.A. et al. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nat. Genet. 43, 491–498 (2011).
20. Anamthawat-Jónsson, K. & Thór Thórsson, A. Natural hybridisation in birch: triploid hybrids between Betula nana and B. pubescens. Plant Cell Tissue Organ Cult. 75, 99–107 (2003).
21. Eidesen, P.B., Alsos, I.G. & Brochmann, C. Comparative analyses of plastid and AFLP data suggest different colonization history and asymmetric hybridization between Betula pubescens and B. nana. Mol. Ecol. 24, 3993–4009 (2015).
22. Patterson, N. et al. Ancient admixture in human history. Genetics 192, 1065–1093 (2012).
23. Ashburner, K., McAllister, H.A. & Rix, M. The Genus Betula: A Taxonomic Revision of Birches (Royal Botanic Gardens, 2013).
24. Wang, N., McAllister, H.A., Bartlett, P.R. & Buggs, R.J.A. Molecular phylogeny and genome size evolution of the genus Betula (Betulaceae). Ann. Bot. 117, 1023–1035 (2016).
25. Lascoux, M., Palmé, A.E., Cheddadi, R. & Latta, R.G. Impact of Ice Ages on the genetic structure of trees and shrubs. Phil. Trans. R. Soc. Lond. B 359, 197–207 (2004).
26. Sofiev, M., Siljamo, P., Ranta, H. & Rantio-Lehtimäki, A. Towards numerical forecasting of long-range air transport of birch pollen: theoretical considerations and a feasibility study. Int. J. Biometeorol. 50, 392–402 (2006).
27. Hollender, C.A. & Dardick, C. Molecular basis of angiosperm tree architecture. New Phytol. 206, 541–556 (2015).
28. Li, P. et al. LAZY1 controls rice shoot gravitropism through regulating polar auxin transport. Cell Res. 17, 402–410 (2007).
29. Dong, Z. & Jin, W. Pleiotropic effects of ZmLAZY1 on the auxin-mediated responses to gravity and light in maize shoot and inflorescences. Plant Signal. Behav. 8, e27452 (2013).
30. Yoshihara, T., Spalding, E.P. & Iino, M. AtLAZY1 is a signaling component required for gravitropism of the Arabidopsis thaliana inflorescence. Plant J. 74, 267–279 (2013).
31. Wang, J., Street, N.R., Scofield, D.G. & Ingvarsson, P.K. Natural selection and recombination rate variation shape nucleotide polymorphism across the genomes of three related Populus species. Genetics 202, 1185–1200 (2016).
32. Korneliussen, T.S., Albrechtsen, A. & Nielsen, R. ANGSD: Analysis of Next Generation Sequencing Data. BMC Bioinformatics 15, 356 (2014).
33. Liu, X. & Fu, Y.-X. Exploring population size changes using SNP frequency spectra.
Nat. Genet. 47, 555–559 (2015).
34. Lynch, M. Evolution of the mutation rate. Trends Genet. 26, 345–352 (2010).
35. Wolfe, J.A. A paleobotanical interpretation of Tertiary climates in the Northern Hemisphere: data from fossil plants make it possible to reconstruct Tertiary climatic changes, which may be correlated with changes in the inclination of the Earth’s rotational axis. Am. Sci. 66, 694–703 (1978).
36. Head, M.J. & Gibbard, P.L. Formal subdivision of the Quaternary System/Period:
past, present, and future. Quat. Int. 383, 4–35 (2015).
37. Gaudet, P. & Dessimoz, C. in The Gene Ontology Handbook (eds. Dessimoz, C. &
Škunca, N.) 189–205 (Springer, 2017).
38. DeYoung, B.J. et al. The CLAVATA1-related BAM1, BAM2 and BAM3 receptor kinase–like proteins are required for meristem function in Arabidopsis. Plant J. 45, 1–16 (2006).
39. Wang, J. et al. The Arabidopsis LRR-RLK, PXC1, is a regulator of secondary wall formation correlated with the TDIF-PXY/TDR-WOX4 signaling pathway. BMC Plant Biol. 13, 94 (2013).
40. Gursanscky, N.R. et al. MOL1 is required for cambium homeostasis in Arabidopsis.
Plant J. 86, 210–220 (2016).
41. Wang, T. et al. A receptor heteromer mediates the male perception of female attractants in plants. Nature 531, 241–244 (2016).
42. Wang, H. & Wang, H. Phytochrome signaling: time to tighten up the loose ends.
Mol. Plant 8, 540–551 (2015).
43. Bartrina, I. et al. Gain-of-function mutants of the cytokinin receptors AHK2 and AHK3 regulate plant organ size, flowering time and plant longevity. Plant Physiol.
173, 1783–1797 (2017).
44. Merchante, C., Alonso, J.M. & Stepanova, A.N. Ethylene signaling: simple ligand, complex regulation. Curr. Opin. Plant Biol. 16, 554–560 (2013).
45. Schlereth, A. et al. MONOPTEROS controls embryonic root initiation by regulating a mobile transcription factor. Nature 464, 913–916 (2010).
46. Doyle, S.M. et al. An early secretory pathway mediated by GNOM-LIKE 1 and GNOM is essential for basal polarity establishment in Arabidopsis thaliana. Proc. Natl.
Acad. Sci. USA 112, E806–E815 (2015).
47. Kinoshita, A. et al. RPK2 is an essential receptor-like kinase that transmits the CLV3 signal in Arabidopsis. Development 137, 3911–3920 (2010).
48. Ueda, M., Zhang, Z. & Laux, T. Transcriptional activation of Arabidopsis axis patterning genes WOX8/9 links zygote polarity to embryo development. Dev. Cell 20, 264–270 (2011).
49. Guan, Y. et al. Phosphorylation of a WRKY transcription factor by MAPKs is required for pollen development and function in Arabidopsis. PLoS Genet. 10, e1004384 (2014).
50. Jiang, W. & Yu, D. Arabidopsis WRKY2 transcription factor mediates seed germination and postgermination arrest of development by abscisic acid. BMC Plant Biol. 9, 96 (2009).
51. Legendre, P. & Fortin, M.-J. Comparison of the Mantel test and alternative approaches for detecting complex multivariate relationships in the spatial analysis of genetic data. Mol. Ecol. Resour. 10, 831–844 (2010).
52. Veyres, N. et al. The Arabidopsis sweetie mutant is affected in carbohydrate metabolism and defective in the control of growth, development and senescence.
Plant J. 55, 665–686 (2008).
53. Goyal, K., Walton, L.J. & Tunnacliffe, A. LEA proteins prevent protein aggregation due to water stress. Biochem. J. 388, 151–157 (2005).
54. Sakuma, Y. et al. Functional analysis of an Arabidopsis transcription factor, DREB2A, involved in drought-responsive gene expression. Plant Cell 18, 1292–1309 (2006).
55. Welling, A. & Palva, E.T. Involvement of CBF transcription factors in winter hardiness in birch. Plant Physiol. 147, 1199–1211 (2008).
56. El Refy, A. et al. The Arabidopsis KAKTUS gene encodes a HECT protein and controls the number of endoreduplication cycles. Mol. Genet. Genomics 270, 403–414 (2003).
57. Bensussan, M. et al. Suppression of dwarf and irregular xylem phenotypes generates low-acetylated biomass lines in Arabidopsis. Plant Physiol. 168, 452–463 (2015).
58. Sugimoto-Shirasu, K. & Roberts, K. “Big it up”: endoreduplication and cell-size control in plants. Curr. Opin. Plant Biol. 6, 544–553 (2003).
59. Gegas, V.C. et al. Endopolyploidy as a potential alternative adaptive strategy for Arabidopsis leaf size variation in response to UV-B. J. Exp. Bot. 65, 2757–2766 (2014).
60. Bonawitz, N.D. & Chapple, C. Can genetic engineering of lignin deposition be accomplished without an unacceptable yield penalty? Curr. Opin. Biotechnol. 24, 336–343 (2013).
61. Bonawitz, N.D. et al. Disruption of Mediator rescues the stunted growth of a lignin- deficient Arabidopsis mutant. Nature 509, 376–380 (2014).
62. Imura, Y. et al. CRYPTIC PRECOCIOUS/MED12 is a novel flowering regulator with multiple target steps in Arabidopsis. Plant Cell Physiol. 53, 287–303 (2012).
63. Mähönen, A.P. et al. Cytokinin signaling and its inhibitor AHP6 regulate cell fate during vascular development. Science 311, 94–98 (2006).
64. Bishopp, A. et al. A mutually inhibitory interaction between auxin and cytokinin specifies vascular pattern in roots. Curr. Biol. 21, 917–926 (2011).
65. Immanen, J. et al. Cytokinin and auxin display distinct but interconnected distribution and signaling profiles to stimulate cambial activity. Curr. Biol. 26, 1990–1997 (2016).
66. Dello Ioio, R. et al. A genetic framework for the control of cell division and differentiation in the root meristem. Science 322, 1380–1384 (2008).
67. Zhu, J. et al. Low temperature inhibits root growth by reducing auxin accumulation via ARR1/12. Plant Cell Physiol. 56, 727–736 (2015).
68. Nguyen, K.H. et al. Arabidopsis type B cytokinin response regulators ARR1, ARR10, and ARR12 negatively regulate plant responses to drought. Proc. Natl. Acad. Sci.
USA 113, 3090–3095 (2016).
69. Lin, R. & Wang, H. Arabidopsis FHY3/FAR1 gene family and distinct roles of its members in light control of Arabidopsis development. Plant Physiol. 136, 4010–4022 (2004).
70. Edwards, K.D., Guerineau, F., Devlin, P.F. & Millar, A.J. Low-temperature-specific effects of PHYTOCHROME C on the circadian clock in Arabidopsis suggest that PHYC underlies natural variation in biological timing. Preprint at bioRxiv https://doi.
org/10.1101/030577 (2015).
© 2017 Nature America, Inc., part of Springer Nature. All rights reserved.
